Coda · Dataset guide Beta Node guide →

Which connectome is right for you?

Coda preconfigures* eleven datasets for you. They differ by specimen, by sex, by which part(s) of the nervous system they cover, by how much of it has been reconstructed/proofread and by what kind of data they come with. You should choose the one that best fits your question**.

Heads-up: this is an opinionated summary.

*You can always bring in your own datasets using the custom CAVE/neuPrint/CATMAID nodes.

**That said, confirming a result in multiple datasets is generally a good idea. Coda provides the tools to do so!

At a glance

A quick overview of the eleven preconfigured datasets. Click on the names to see details.

DatasetSpecimenRegionNeuronsCompletenessEM resolutionYear
MaleCNSAdult male DrosophilaWhole CNS — brain and ventral nerve cord~166,700 proofreadDense8 × 8 × 8 nm2026
FlyWire FAFB publicAdult female DrosophilaWhole brain (central brain + both optic lobes)~140,000Dense4 × 4 × 40 nm2024
MICrONS Minnie65 publicMouseVisual cortex, ~1 mm³~75,000 with reconstructed morphologyDense segmentation, partial proofreading8 × 8 × 40 nm2025
HemibrainAdult female DrosophilaCentral brain, roughly one hemisphere~25,000 traced (some truncated)Dense within the volume8 × 8 × 8 nm2020
MANCAdult male DrosophilaVentral nerve cord~15,800Dense8 × 8 × 8 nm2024
BANC publicAdult female DrosophilaWhole CNS — brain and ventral nerve cord~142,000Dense4 × 4 × 45 nm2026
L1First-instar Drosophila larvaWhole CNS — brain to abdominal neuromeres5,013 hand-tracedDense in the brain, sparse elsewhere3.8 × 3.8 × 50 nm2015 (volume), 2023 (brain connectome)
FAFBAdult female DrosophilaWhole brain (volume); reconstructions are sparseA few thousand hand-tracedSparse — traced to answer specific questions4 × 4 × 40 nm2018 (volume) + subsequent papers
Optic LobeAdult male DrosophilaRight optic lobe — medulla, lobula, lobula plate, part of the lamina~50,000Dense8 × 8 × 8 nm2025
FIB-19Adult female DrosophilaPortions of medulla, lobula and lobula plateHundredsDense within a small columnar volume8 × 8 × 8 nm2017
Mushroom BodyAdult male DrosophilaMushroom body alpha (vertical) lobe983Dense within the lobe8 × 8 × 8 nm2017

Demo Data is not in the table: it is a synthetic dataset for demonstration purposes only. It needs no account and fetches nothing.

Start here

If you have no reason to pick otherwise, pick one of these. They are our first picks in their respective category.

MaleCNS

The whole central nervous system of an adult male fly.

Specimen
Adult male Drosophila
Region
Whole CNS — brain and ventral nerve cord
Neurons
~166,700 proofread
Completeness
Dense
EM resolution
8 × 8 × 8 nm
Published
2026

A whole-CNS (brain and nerve cord) reconstruction. A collaboration between the Janelia FlyEM project, Google Connectomics and the University of Cambridge. The largest and most complete fly connectome to date.

Pros

  • High synaptic completion rate (fraction of synapses attached to proofread neurons)
  • Brain & nerve cord in one volume lets you trace full circuits from sensory input to motor output
  • Densely annotated with e.g. types, classes, hemilineages, neurotransmitters, sexual dimorphism and cross-referenced to the literature
  • Cross-references to hemibrain, FlyWire and MANC as first-class columns
  • Served via neuPrint, which provides the most complete set of queries

Cons

  • None, unless you need a female brain

Where it lives

  • neuPrintOpens in Coda as MaleCNS (neuPrint)connectivity, annotations and skeletons; needs a free account for a token
  • Project pageproject overview with links to raw data downloads

How to cite it

Working with it outside Coda

FlyWire FAFB public

A whole female brain, optic lobes included.

Specimen
Adult female Drosophila
Region
Whole brain (central brain + both optic lobes)
Neurons
~140,000
Completeness
Dense
EM resolution
4 × 4 × 40 nm
Published
2024

An automated segmentation of the FAFB volume (Zheng et al., 2018), proofread by Princeton, Cambridge and the FlyWire community. Published with a hierarchical annotation set covering super class, class, cell type and side.

Coda reads it through CAVE, so you pick a version by its materialization number: 783 matches the Nature paper package and 630 the preprint.

The annotation table inside the datastack is out of date. If you start a FlyWire workflow from the New menu or the Workflow Wizard, Coda therefore adds the current annotations in front of the dataset: the hierarchical annotations from the flywire_annotations repository (with root ids updated) plus the community annotations as tags.

Pros

  • A whole brain including both optic lobes
  • Dense annotations from two sources: the hierarchical set & community tags
  • High synaptic completion rate (fraction of synapses attached to proofread neurons)

Cons

  • Brain only, nothing below the neck
  • CAVE does not support certain queries, e.g. paths or per-region connection counts
  • High-res skeletons exist only on materialization 783

Where it lives

  • Codexthe published browser: cell type search, circuit summaries, downloads and more
  • CAVEOpens in Coda as FlyWire FAFB publicthe datastack Coda queries; segmentation, synapses and meshes
  • flywire_annotationsthe current hierarchical annotations, which Coda adds to new FlyWire workflows
  • Nature paper packagelanding page for the FlyWire paper package

How to cite it

Working with it outside Coda

Pick this over MaleCNS when you need the female brain. If you need a female brain + nerve cord, take a look at the BANC dataset.

Open it in a workflow

MICrONS Minnie65 public

A cubic millimetre of mouse visual cortex, with functional recordings from the same neurons.

Specimen
Mouse
Region
Visual cortex, ~1 mm³
Neurons
~75,000 with reconstructed morphology
Completeness
Dense segmentation, partial proofreading
EM resolution
8 × 8 × 40 nm
Published
2025

The MICrONS collaboration's cortical volume: a dense EM segmentation of mouse visual cortex, registered to two-photon calcium recordings of the same neurons responding to visual stimuli.

Pros

  • Mammalian cortex at synaptic resolution
  • Structure and function for the same neurons

Cons

  • Proofreading is partial (but ongoing): only a subset of neurons are complete, so connectivity is a lower bound and axons are truncated
  • A cubic millimetre is a small fraction of one cortical area, so long-range connections are missing
  • Annotations are coarse compared with anything in the fly

Where it lives

How to cite it

Working with it outside Coda

Alternatives

The datasets below all have their place but some come with caveats you should be aware of.

Hemibrain

The first large-scale dense reconstruction of a fly brain, and the template for a lot of subsequent work.

Specimen
Adult female Drosophila
Region
Central brain, roughly one hemisphere
Neurons
~25,000 traced (some truncated)
Completeness
Dense within the volume
EM resolution
8 × 8 × 8 nm
Published
2020

Janelia FlyEM's reconstruction of approximately half the central brain. It extends across the midline to include the central complex. Most later work uses its cell type names.

Pros

  • Probably the best-proofread mushroom body and central complex

Cons

  • Half a brain, so no bilateral symmetry to check results against
  • Neurons are truncated at the midline and at the optic lobe boundary, which needs to be taken into account when analysing connectivity, cell counts, morphology, etc.

Where it lives

How to cite it

Working with it outside Coda

MANC

The ventral nerve cord of an adult male fly: motor neurons, premotor circuits and the descending neurons that drive them.

Specimen
Adult male Drosophila
Region
Ventral nerve cord
Neurons
~15,800
Completeness
Dense
EM resolution
8 × 8 × 8 nm
Published
2024

A dense reconstruction of the male ventral nerve cord from Janelia FlyEM, the Cambridge Drosophila Connectomics Group and Google Connectomics. Motor neurons, the premotor networks driving them, and the descending neurons arriving from the brain are all typed and systematically named.

Pros

  • Dense, systematic annotations
  • High completion rate (fraction of synapses attached to proofread neurons)

Cons

  • Nerve cord only: descending/ascending neurons are cut at the neck
  • Superseded by MaleCNS for new work

Where it lives

How to cite it

Working with it outside Coda

Reach for this dataset if you want to focus on the ventral nerve cord alone or to confirm results from e.g. the MaleCNS.

Open it in a workflow

BANC public

The whole central nervous system of an adult female fly, with the neck connective intact. The female counterpart to MaleCNS.

Specimen
Adult female Drosophila
Region
Whole CNS — brain and ventral nerve cord
Neurons
~142,000
Completeness
Dense
EM resolution
4 × 4 × 45 nm
Published
2026

Brain And Nerve Cord: a complete female central nervous system imaged in one piece, with detailed annotation of the neurons innervating sensory organs, the motor neurons, and the viscera.

Like FlyWire, this is a CAVE datastack, which limits some of the analysis capabilities.

Pros

  • Whole CNS of a female fly, so you can compare it against MaleCNS to look for sex differences
  • Intact neck connective, so descending and ascending neurons are complete
  • Strong peripheral annotation (sensory modality, motor neurons, muscles, biological context, etc.)

Cons

  • Completion rate (fraction of synapses attached to proofread neurons) is generally much lower than in MaleCNS, hemibrain or FlyWire
  • At the time of writing there are still neurons being proofread in the production datastack
  • CAVE datastacks do not support certain queries, e.g. paths or per-region connection counts

Where it lives

  • Codexthe primary browser: cell type search, circuit summaries, downloads and more
  • Project overviewthe landing page for the BANC project

How to cite it

Working with it outside Coda

L1

The first-instar larval central nervous system.

Specimen
First-instar Drosophila larva
Region
Whole CNS — brain to abdominal neuromeres
Neurons
5,013 hand-traced
Completeness
Dense in the brain, sparse elsewhere
EM resolution
3.8 × 3.8 × 50 nm
Published
2015 (volume), 2023 (brain connectome)

A whole first-instar larval CNS, hand-traced in CATMAID and hosted by Virtual Fly Brain. The brain was reconstructed to completion and published as a full connectome; the rest (SEZ, nerve cord) is traced sparsely.

Pros

  • A complete brain connectome of a behaving animal
  • Rich behavioural and genetic literature tied to identified neurons

Cons

  • Sparse outside the brain

Where it lives

How to cite it

Working with it outside Coda

Legacy datasets

These predate the more recent dense, large-scale datasets. Kept for historical reference and as playgrounds.

FAFB

Manually traced neurons in the EM volume that eventually became FlyWire.

Specimen
Adult female Drosophila
Region
Whole brain (volume); reconstructions are sparse
Neurons
A few thousand hand-traced
Completeness
Sparse — traced to answer specific questions
EM resolution
4 × 4 × 40 nm
Published
2018 (volume) + subsequent papers

The Female Adult Fly Brain volume, with sparse manual reconstructions in CATMAID from many labs over several years. Coda reads the published subset hosted by Virtual Fly Brain.

Note that while this is the same EM volume as FlyWire, the latter is a re-aligned version of the former, so coordinates will differ somewhat.

Pros

  • Some very high-quality hand-traced neurons (mileage may vary though)

Cons

  • Sparse: most of the brain was never traced, so a missing partner does not mean there is no connection

Where it lives

How to cite it

Working with it outside Coda

This is mostly legacy data at this point. Unless you have very specific reasons to use this, you should reach for FlyWire instead.

Open it in a workflow

Optic Lobe

One optic lobe of the MaleCNS volume, released ahead of the whole thing.

Specimen
Adult male Drosophila
Region
Right optic lobe — medulla, lobula, lobula plate, part of the lamina
Neurons
~50,000
Completeness
Dense
EM resolution
8 × 8 × 8 nm
Published
2025

The right optic lobe of the MaleCNS volume, with parts of the central brain and lamina, published separately and earlier. It contains the visual system cell type inventory used by the optic lobe papers.

The same tissue is also part of MaleCNS. Use this release if you want the inventory exactly as published.

Pros

  • The published visual cell type inventory, with the naming the optic lobe literature uses
  • Much smaller than MaleCNS, so optic lobe analyses run faster

Cons

  • A subset of MaleCNS: anything leaving the optic lobe is cut
  • Kept mainly as a reference for the papers that used it

Where it lives

  • neuPrintOpens in Coda as Optic Lobe (neuPrint)

How to cite it

Working with it outside Coda

FIB-19

An early partial reconstruction of the female visual system, made to study motion detection.

Specimen
Adult female Drosophila
Region
Portions of medulla, lobula and lobula plate
Neurons
Hundreds
Completeness
Dense within a small columnar volume
EM resolution
8 × 8 × 8 nm
Published
2017

A focused ion beam reconstruction covering a few columns of the fly visual system, made to map the circuits behind elementary motion detection.

It has been superseded by the Optic Lobe release and MaleCNS. We keep it because much of the motion detection literature is based on it.

Pros

  • The volume the ON/OFF motion detection circuit papers were built on

Cons

  • Covers only a few columns, so check any result against a complete volume before generalising

Where it lives

  • neuPrintOpens in Coda as FIB-19 (neuPrint)

How to cite it

Mushroom Body

A dense reconstruction of the mushroom body alpha lobe, and one of the first fly connectomes.

Specimen
Adult male Drosophila
Region
Mushroom body alpha (vertical) lobe
Neurons
983
Completeness
Dense within the lobe
EM resolution
8 × 8 × 8 nm
Published
2017

A dense reconstruction of the alpha lobe, published as one of the first demonstrations that a learning and memory centre could be mapped at synaptic resolution.

It contains 983 neurons. We include it because the mushroom body learning literature cites it.

Pros

  • Small (983 neurons) and completely reconstructed within its boundary

Cons

  • Only one lobe of the mushroom body: every input and output is cut at the boundary
  • Both hemibrain and MaleCNS contain the whole mushroom body

Where it lives

  • neuPrintOpens in Coda as Mushroom Body (neuPrint)

How to cite it

What else is out there

These connectomes have no dedicated node in Coda, but you can still open most of them: the Custom CAVE, Custom neuPrint and Custom CATMAID nodes take a server and a dataset name, and the Neuroglancer Source node takes a precomputed bucket.